The term machine learning (ML) was first named in 1959, though it has certainly grown in prominence and effectiveness in successive decades. In today’s business world, it’s a trendy buzzword that isn’t fully understood by many.
A unique subset of artificial intelligence (AI), ML allows computer systems to analyze necessary data and learn based on outcomes, using models, patterns, and little outside intervention. ML has such widespread applications that most consumers aren’t even aware that it’s happening behind the scenes. The adaptability and reliability of the technology is being harnessed in industries ranging from social media to self-driving cars.
How one industry expertly leverages ML
In fintech, one of the first—and still the most prominent—examples of ML is fraud detection, particularly in credit cards. End users and credit card companies both benefit from being able to identify fraud quickly and accurately. For years, these groups have employed ML to gather and analyze data including transaction location, types of purchases, and merchant categories with the goal of identifying patterns and flagging potential anomalies
By using ML to improve the speed and accuracy of fraud detection, an entire industry has drastically reduced its dependence on consumer input and shortened the timeframe that a given account might be misused, thereby reducing costs drastically. Credit, debit and prepaid card businesses saw nearly $22 billion in worldwide fraud in 2015; and with fraud on the rise, ML and other approaches to big data are playing an essential role in controlling risk. And its very nature has made ML especially effective in this pursuit. As the perpetrators of fraud shift tactics, ML systems learn and adapt, helping to keep pace with change dynamically.
How machine learning helps small businesses
In the small business lending space, online lenders and marketplaces like Lendio are using ML to simplify the loan application process for small business owners and shorten the timeframe for receiving financing.
"A unique subset of artificial intelligence (AI), ML allows computer systems to analyze necessary data and learn based on outcomes, using models, patterns, and little outside intervention."
In the case of Lendio’s marketplace model, ML works to analyze relevant data and match a business owner with the best potential lenders among a mix of diverse loan product offerings. Rather than go from bank to bank filling out loan applications, this technology allows small business owners to fill out one online application and receive loan offers that match their qualifications and needs. The whole process can be done in as little as 24 hours.
The ease and speed at which ML can take in data and share recommendations has made it a boon for businesses such as Porters Bar & Grill in Boston, a popular destination for Celtics and Bruins fans and other downtown Boston event goers. The bar struggled with cash flow during the off season and faced potential closure. At that challenging time, Porters’ owners had been turned down for a loan by their own bank, and didn’t have dozens of hours to explore banks and other non-traditional funding options to find one that would fit their specific needs. Instead, Lendio’s ML identified two sources of funds that had a greater-than-50-percent chance of working with this specific loan request (while most others were well below 20 percent). Thanks to ML, Porters found the best loan for its unique financial needs, and quickly. This allowed the business to remain open and position itself to earn over $1MM in annual revenue.
See Also: Top Fintech Solution Companies In Europe
How consumers benefit
Outside of the business lending realm, money management services like Mint use ML to boost consumers’ financial well-being. By aggregating spending data and analyzing charges, the company can recommend ways to save money. Whether that’s a mortgage refinance or stronger investment opportunities, users gain invaluable spending insight. Other entities, such as credit card companies and banks will use ML to automatically classify expenses and aid consumers in budgeting.
While it can’t replace the human touch that’s important in many financial services, ML can assist with customer service. By using chatbots, many companies have streamlined their customer interactions and sped the process of connecting people with the information they’re seeking.
How to get everyone else on board
While a CIO might readily appreciate the benefits of ML in a given application, getting buy-in from other decision-makers isn’t a given. The technology may not link directly to a lift in sales, and it will take time to recognize the full weight of its impact. Because many stakeholders struggle to understand how ML works, internally, it’s important to help executives understand the applications of ML in a given business, while clearly communicating the value it can provide.
Executives seeking to invest in ML must also be able to prepare stakeholders for the potential shortcomings of the route. The technology will make mistakes. This is part of the process of learning and improving accuracy. If they’re not prepared to expect mistakes, some team members may be inclined to hastily throw the whole system out based on a few misses.
In a speech at TechCrunch Disrupt SF, John Giannandrea, senior vice president of engineering at Google, said, “machine learning and artificial intelligence is extremely important and will revolutionize many vertical industries.” As it continues to improve the way that companies operate, leaders in all industries should take notice and consider how ML could work for them. The potential applications of ML in the fintech space are still being explored, but businesses that overlook the technology do so at their own potential peril.